[ayewo]

Here's a recent comment [1] by an OpenAI engineer confirming that they do in fact make such trade offs between intelligence and efficiency.

[1]: https://news.ycombinator.com/item?id=46909905

[int_19h]

That comment only says that they have a lot of different options for smaller & faster models that people can opt into. It doesn't say that they dynamically scale things up or down depending on demand.